Step 1: Develop Unit Process Templates Using Plant Block Diagrams

"Let's discuss how we build a digital plant model," said Peter. Turton et al. (2018) describes best practices for building process block and flow diagrams. Updated process block, flow, and piping and instrumentation diagrams, used with real-time data, achieve best practices for process monitoring (Bascur 2019).

Online, real-time configurable calculations can be created using EIDI analytics, with event frames created to monitor the exact state of the process and its duration This capability enables the team to use an EIDI template to generate operational modes for all units. These operational modes are then used to set the event framing start and end times, which generates the minimum, maximum, total, and average and standard deviation value for each variable in the event template. The ability to integrate data from multiple sources and

FIGURE 4.3

Oil refinery block diagram transformed to a digital plant schema. (Courtesy of O.A. Bascur, OSIsoft LLC.)

apply logic combining real time and external channels offers a variety of methods for determining KPIs.

Our journey starts by configuring unit templates with live process variables. Let's take a look at our refinery's process block diagram, which is shown in Figure 4.3.

Figure 4.3 shows the process block diagram in relation to the hierarchical structures represented by a single, reusable unit template. Peter explained that the unit template is part of the EIDI and the concept has been used in other process industries for many years.

The transition from calendar- to condition-based maintenance provides the ability to capture information about various states of equipment. For example, determining the number of starts and stops over time can help to determine appropriate maintenance cycles. The ability to capture these events along with related information such as leading indicators can provide valuable insight into the prevention of future failure.

Each time a unit changes its operational state, it records an event frame. The event frame subsystem marks the time intervals for further processing and analysis. The time intervals determine aggregation of actual production and consumable data values, to estimate production and consumption losses (totals, averages, standard deviation, minimum and maximum).

The EIDI contains many calculations and algorithms to handle time-series data, such as interpolation of sampled laboratory data with real-time production data to group them together when developing process models (Steyn et al. 2018; Bascur and Soudek 2019).

The EIDI unit template event frame captures process unit rates and consumables to automatically calculate average production rate, total energy, and water and air consumed for each operating mode. Table 4.1 shows the typical results.

Event Frame Data Created for Extraction and Analysis

Event Frame

Asset

Start

End

Duration

Mode

Process Feed Rate

Electricity

Consumption

Water

Consumption

Other

Variables

Analysis

template

20120725

Boiler

August 1, 2019

12:44:00 p.m.

August 2, 2019

3:55:00 a.m.

15:01:00

Running

131.5

84.4

20.5

XX.X

Trouble state duration

Fluid catalytic cracking unit (FCCU)

August 2, 2019

3:55:00 a.m.

August 2, 2019 4:00:00

00:05:00

Trouble

0.0

30.0

10

X.XX

Down state time

Desalter

August 2, 2019

August 2, 2019

00:12:00

Down

0.0

15.0

5.0

X.XX

Running OK state duration

Alkylation

August 2, 2019

August 2, 2019

03:46:00

Running

95.0

78.0

45.0

X.XX

Maintenance

state

duration

Vacuum

tower

August 2, 2019

August 2, 2019

00:10:00

In maintenance

0.0

20.0

5.0

X.XX

FIGURE 4.4

Smart process unit template components.

Monica explained further, "We can derive the operational status by monitoring real-time production targets and comparing actual results. Each unit calculation is identical, creating event frames to indicate various operational modes." Figure 4.4 shows the basic building blocks of a process unit template, which can be replicated for other units and refineries.

Monica offered, "The same production unit template can be reused at other Proclndustries refineries to help them get started when implementing their EIDI. It's best to start simple and not reinvent the wheel."

Figure 4.5 shows Proclndustries' production targets and process variables, which are the inputs to the process unit template. Monica pointed out the system side of real-time intelligence that uses powerful analytics and eventframing capabilities: "Are we on target?" "Are we satisfied?" represents the human side, providing information and operational insights for further decision-making (Bascur and Halhead 2013).

A unique data model is the skeleton for all areas of the process plant. This data model is like DNA; it is an object model that has the key attributes that describe the performance of a process unit. In this case, a process unit takes a process feed rate with reagents, energy, fuel, water, and air to generate product stream(s).

Monica explained that they use production rate, consumable electricity, and water and air totals in the unit template and five operating state definitions: "running OK," "trouble," "idle," "down," or "in maintenance." During production runs, this modular, reusable unit template calculates and segments time into one of those operational modes.

FIGURE 4.5

Smart unit template to evaluate overall refinery production performance. (Courtesy of O.A. Bascur, OSIsoft LLC.)

 
Source
< Prev   CONTENTS   Source   Next >